COMPLAS 2023

Robust topology optimisation of lattice structures with spatially correlated imperfections

  • Ben-Yelun, Ismael (Universidad Politécnica de Madrid)
  • Yuksel, Ahmet Oguzhan (University of Cambridge)
  • Cirak, Fehmi (University of Cambridge)

Please login to view abstract download link

The member stiffnesses and thicknesses of additively manufactured lattices are subject to spatially and directionally correlated anisotropic uncertainties, which must be considered in assessing and optimising their structural performance. We propose a novel topology optimisation approach for the robust structural design of lattice structures in which the member cross-sectional areas are the design variables. Structural compliance is minimised while taking into account the inherent uncertainties resulting from the manufacturing process. A physically derived random field representation is used to model isotropic or anisotropic and homogenous or non-homogeneous fields. For robust optimisation, a standard gradient-based First-Order Second-Moment (FOSM) approach has been adopted and solved with the Method of Moving Asymptotes (MMA). The versatility of the proposed method is demonstrated through several numerical examples, from illustrative to more complex 2D and 3D problems.